AI Paper Atlas

Research map / Detection and segmentation / CNN features

Deep Direct Regression for Multi-Oriented Scene Text Detection

Wenhao He, Xu-Yao Zhang, Fei Yin, Cheng-Lin Liu

ICCV 2017 · Detection and segmentation › CNN features · approach: Arbitrary-shape scene text detection · 426 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use Recalled NMS instead of traditional NMS as post-processing (ablation)Proposed direct regression detector · ICDAR2015 Incidental Scene TextF-measure: 0.80 -> 0.81Improved
use Recalled NMS instead of traditional NMS as post-processing (ablation)Proposed direct regression detector · ICDAR2015 Incidental Scene Textprecision: 0.81 -> 0.82Improved
use direct regression instead of indirect (proposal-based) regressionmulti-oriented text detector · ICDAR2015 Incidental Scene TextF-measure: 0.71 -> 0.81Improved
train without MSRA-TD500 training set or Chinese data (English word-level only)Proposed direct regression detector · MSRA-TD500F-measure: 0.74 vs 0.74 (Zhang et al.)No clear effect
direct regression detector versus indirect-regression-free SSD/Faster-RCNN style baseline on speedProposed direct regression detector · ICDAR2013 Focused Scene Texttime per image: 0.9s vs 0.73s (Liao et al.), 2.1s (Zhang et al.)Mixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy